There is growing evidence that certain facets of heart disease may be influenced by infections arising from the oral cavity. Streptococcus mutans, a major etiologic agent of dental caries, is also a leading causative agent of infective endocarditis (IE) and has been implicated in the development and progression of atherosclerosis. We have recently demonstrated that certain S. mutans strains invade and persist in the cytoplasm of Human Coronary Artery Endothelial Cells (HCAEC). Invasion of HCAEC was associated with an ability to avidly bind to collagen and laminin, and invasive strains were significantly more virulet than non-invasive strains in the Galleria mellonella model for systemic infection. A collagen-binding protein (Cnm) was found only in invasive strains, and inactivation of the cnm gene significantly impaired collagen and laminin binding capacity, abolished HCAEC invasion, and attenuated virulence in G. mellonella. Characterization of a gene encoding a putative glycosyltransferase (CsbB) co-transcribed with cnm suggested that CsbB is involved in Cnm glycosylation. Preliminary data using a rabbit endocarditis model suggested that expression of Cnm enables infection of the underlying endothelium by S. mutans whereas a cnm mutant was trapped in the heart valve vegetation. Collectively, our data suggest that the ability to invade an persist in the host cytoplasm is an important, previously unrecognized, virulence trait of S. mutans. Our working hypothesis is that S. mutans strains expressing Cnm have an enhanced capacity to colonize human tissues, and that this ability is a major virulence trait in systemic infections. The major goals of this application are to conduct the functional analysis of Cnm and to characterize the role of Cnm in host-bacteria interactions. To achieve these goals, we propose three well-integrated Specific Aims. In Aim 1 (Molecular characterization of the cnm locus), we will (i) conduct the transcriptional characterization of the cnm operon, (ii) employ mutagenesis approaches to continue the characterization of the Cnm and CsbB, and (iii) identify the essential machinery for intracellular invasion using an heterologous host system. In Aim 2 (Functional analysis of Cnm), we will (i) use recombinant variant forms of Cnm to identify the functional domains of Cnm, and (ii) determine whether Cnm is glycosylated and, if so, how Cnm is modified. In Aim 3 (Streptococcus mutans-host interactions), we will (i) use in vitro models to assess Cnm-platelet interactions, (ii) use an ex vivo adhesion model to examine the ability of invasive and non-invasive strains to colonize heart tissues, and (iii) use a rabbit endocarditis model to establish the contribution of Cnm to the pathogenesis of IE and to test the efficacy of active immunization with Cnm to prevent S. mutans infections. This study has strong potential to be readily translated into clinical settings as Cnm could serve as a biomarker for detection of hypervirulent strains in patients in need to receive prophylactic treatment, or as a potential target for the development of effective strategies for prevention and treatment of S. mutans extra-oral infections.